Over 5.5 million Americans over 65 have Alzheimer?s Disease (AD) and Alzheimer?s Disease-related dementias (ADRD), and many of them have multiple coexisting chronic conditions. Decision-making for chronic conditions often defaults to disease-specific guidelines, which may be inappropriate in the setting of AD/ADRD. Clinical trials often exclude older adults with AD/ADRD, and the benefits and harms of treatments may be quite different than in younger, healthier individuals due to declines in life expectancy, cognition and function. Importantly, guidelines may not address the variable outcomes that patients value most. The decision regarding whether or not to prescribe anticoagulation for a patient with atrial fibrillation is an exemplar of a complex clinical decision often made in the context of coexisting AD/ADRD. Atrial fibrillation is the most common arrhythmia in older adults, and places those affected at significantly increased stroke risk. The current decisional paradigm regarding anticoagulation relies on weighing the risk of stroke versus the risk of bleeding. By virtue of their age, nearly all patients with atrial fibrillation and AD/ADRD have stroke risk scores that meet the threshold for prescription of an oral anticoagulant. However, the appropriateness of anticoagulation may change with shifting benefits and harms and goals of care over the course of AD/ADRD. Little is currently known about the factors involved in clinical decision-making about anticoagulation for atrial fibrillation in the context of comorbid AD/ADRD. Although a few studies have demonstrated that an AD/ADRD diagnosis is associated with lower levels of anticoagulation, there is a paucity of evidence about how dementia progression influences decisions to continue or discontinue anticoagulant prescription. There has also been little work to understand how clinicians ultimately make these decisions. The proposed work will redress these gaps with a mixed-methods approach to understand the decisions made by clinicians regarding anticoagulation for patients with atrial fibrillation and comorbid AD/ADRD, and how they arrived at these decisions. Contemporary data from the electronic medical record (EMR) of a large dementia care practice will be leveraged to understand functional, clinical, and sociodemographic factors associated with anticoagulant use and discontinuation over the course of AD/ADRD progression. This work will also employ rigorous qualitative methods to capture the range of factors physicians consider and how they integrate these factors when making decisions about anticoagulation for this vulnerable population. This work represents a first step in a research career devoted to promoting optimal decision-making about comorbid chronic conditions for patients with Alzheimer?s Disease and Alzheimer?s Disease-related dementias. This GEMSSTAR Award would allow the candidate, a geriatrician, to obtain the training and mentorship to further his quantitative and qualitative analysis skills and develop expertise in using EMR data to pursue research at the forefront of clinical geriatrics.